Research On Multi-state Joint Estimation Of Power Lithium Battery Based On MIAUKF Algorithm

Lithium battery is one of the power sources of electric vehicles.Its low energy efficiency and mileage anxiety problems restrict the development of electric vehicles.BMS is mainly for effective battery management and advance warning to ensure safety and stability operation of electric vehicles.Important parameters predicted by the battery management system include state of charge(SOC)and state of health(SOH).They are crucial points to determine whether the battery is properly charged and discharged,whether the electric vehicle can drive safely and the remaining mileage distance.Contraposing to the major factors affecting the state estimation accuracy of lithium battery,examples include the selection of circuit model,the identification of model parameters,the filtering method and noise influences.This text regards INR18650-30 Q single cell as the study object and carries out a series of research work:(1)The experiment platform is constructed.Lithium battery characteristic test and dynamic condition experiment are designed respectively,which establish a data foundation for subsequent model parameter identification and state estimation.(2)Comprehensive consideration of virtues and defects of various types equivalent circuit models,the second-order RC model is chosen as the model foundation in this paper.Establish state space equation and observation equation.The model parameter values are identified offline as reference values when SOC = 0.9,and the high accuracy of the model is verified under HPPC cycle condition.(3)Aiming at the problem that the model parameters are fixed and not updated with the change of working conditions and environment,the variable forgetting factor multi-innovation least squares method is proposed to identify the model parameters online and the parameter variation curve is given.The results show that the online parameter identification method can on-line modification and real-time update the model parameters within the allowable deviation range of the initial values of the model parameters.(4)The basic Kalman series algorithms are introduced.Focusing on the deficiencies of EKF and UKF algorithms,AUKF algorithm as the foundation and combining with multi-innovation identification theory,MIAUKF algorithm is put forward to estimate the SOC of lithium battery.The precision and adaptability of the improved algorithm are verified by comparison under various working conditions and algorithms.(5)Aiming at the problem that the algorithm that only estimates SOC does not adapt to the capacity change caused by battery aging,which leads to the increase of estimation error,the VFFMILS-MIAUKF joint algorithm is proposed for the joint estimation of state of charge and state of health of lithium battery,and SOC and SOH joint estimation under four working conditions is carried out.Through the error curve and error analysis table compared to verify the meliority of the joint algorithm.At the same time,the influence of different SOC initial values is analyzed,which further proves the suitability and robustness of the joint algorithm.Figure [61] table [7] reference [64]...